Abstract

Atmospheric blocking is a large-scale quasi-stationary phenomenon in mid-latitude circulation, characterized by persistent high-pressure systems that disrupt the typical west-to-east flow of the jet stream. These systems can cause extreme weather events—such as heatwaves, cold spells, or droughts—that persist for days or even weeks. This study proposes a deep learning framework to predict and interpret the occurrence of atmospheric blocking by integrating geophysical precursors such as geopotential height (Z500), stream function (SF200), and potential vorticity. These features, which are dynamically linked to blocking onset and persistence, serve as inputs to a Convolutional Neural Network model trained on the CESM Large Ensemble (LENS) dataset. After initial training, we apply transfer learning to adapt the model to the ERA5 reanalysis dataset, preserving the feature extractor while retraining the classifier to improve generalization on real-world atmospheric data. To improve model transparency and interpretability, we apply eXplainable AI techniques. These tools provide spatial insights into the CNN's decision-making process, highlighting which atmospheric features are most important for the model’s blocking predictions. By advancing the understanding of mechanisms and indicators of blocking through deep learning and XAI, this work contributes toward more accurate and interpretable weather forecasting models—vital for anticipating and mitigating the societal impacts of extreme weather events.

Keywords

Atmospheric Blocking, Deep Learning, Convolutional Neural Network

Date of this Version

8-5-2025

Share

COinS